Although bioactive peptides can be produced chemically by a variety of synthesis strategies, recombinant technology offers the potential for inexpensive, large-scale production of peptides without the use of organic solvents, highly reactive reagents or potentially toxic chemicals. However, expression of short peptides in Escherichia coli and other microbial systems can sometimes be problematic. For example, short peptides are often degraded by the proteolytic and metabolic enzymes present in microbial host cells. Use of a fusion protein to carry the peptide of interest may help avoid cellular degradation processes because the leader protein is large enough to protect the peptide from proteolytic cleavage. Moreover, certain fusion proteins can direct the peptide to specific cellular compartments, i.e. cytoplasm, periplasm, inclusion bodies or media, thereby helping to avoid cellular degradation processes. However, while use of a fusion protein may solve certain problems, cleavage and purification of the peptide away from the fusion protein can give rise to a whole new set of problems.
Preparation of a peptide from a fusion protein in pure form requires that the peptide be released and recovered from the fusion protein by some mechanism. In many cases, the peptide of interest forms only a small portion of the fusion protein. For example, many peptidyl moieties are fused with β-galactosidase that has a molecular weight of about 100,000 daltons. A peptide with a molecular weight of about 3000 daltons would only form about 3% of the total mass of the leader protein. Also, separate isolation or purification procedures (e.g. affinity purification procedures) are generally required for each type of peptide released from a fusion protein. Release of the peptide from the fusion protein generally involves use of specific chemical or enzymatic cleavage sites that link the carrier protein to the desired peptide [Forsberg et al., Int. J. Protein Chem., 11: 201-211, (1992)]. Chemical or enzymatic cleavage agents employed for such cleavages generally recognize a specific sequence. However, if that cleavage sequence is present in the peptide of interest, then a different cleavage agent must usually be employed. Use of a complex fusion partner (e.g. β-galactosidase) that may have many cleavage sites produces a complex mixture of products and complicates isolation and purification of the peptide of interest.
Chemical cleavage reagents in general recognize single or paired amino acid residues that may occur at multiple sites along the primary sequence, and therefore may be of limited utility for release of large peptides or protein domains which contain multiple internal recognition sites. However, recognition sites for chemical cleavage can be useful at the junction of short peptides and carrier proteins. Chemical cleavage reagents include cyanogen bromide, which cleaves at methionine residues [Piers et al., Gene, 134: 7, (1993)], N-chloro succinimide [Forsberg et al., Biofactors, 2: 105-112, (1989)] or BNPS-skatole [Knott et al., Eur. J. Biochem., 174: 405-410, (1988); Dykes et al., Eur. J. Biochem., 174: 411-416, (1988)] which cleave at tryptophan residues, dilute acid which cleaves aspartyl-prolyl bonds [Gram et al., Bio/Technology, 12: 1017-1023, (1994); Marcus, Int. J. Peptide Protein Res., 25: 542-546, (1985)], and hydroxylamine which cleaves asparagine-glycine bonds at pH 9.0 [Moks et al., Bio/Technology, 5: 379-382, (1987)]. Forsberg et al., Int. J. Protein Chem., 11: 201-211, (1992)
For example, Shen describes bacterial expression of a fusion protein encoding pro-insulin and β-galactosidase within insoluble inclusion bodies where the inclusion bodies were first isolated and then solubilized with formic acid prior to cleavage with cyanogen bromide. [Shen, Proc. Nat'l. Acad. Sci. (USA), 281: 4627 (1984)]. Dykes et al. describes soluble intracellular expression of a leader protein encoding α-human atrial natriuretic peptide and chloramphenicol acetyltransferase in E. coli where the fusion protein was chemically cleaved with 2-(2-nitrophenylsulphenyl)-methyl-3′-bromoindolenine to release peptide. [Dykes et al., Eur. J. Biochem., 174: 411 (1988)]. Ray et al. describes soluble intracellular expression in E. coli of a fusion protein encoding salmon calcitonin and glutathione-S-transferase where the fusion protein was cleaved with cyanogen bromide. [Ray et al., Bio/Technology, 11: 64 (1993)]
Proteases can provide gentler cleavage conditions and sometimes even greater cleavage specificity than chemical cleavage reagents because a protease will often cleave a specific site defined by the flanking amino acids and the protease can often perform the cleavage under physiological conditions. For example, Schellenberger et al. describes expression of a fusion protein encoding a substance P peptide (11 amino acids) and β-galactosidase within insoluble inclusion bodies, where the inclusion bodies were first isolated and then treated with chymotrypsin to cleave the leader protein. [Schellenberger et al., Int. J. Peptide Protein Res., 41: 326 (1993)]. Pilon et al. describe soluble intracellular expression in E. coli of a leader protein encoding a peptide and ubiquitin where the fusion protein was cleaved with a ubiquitin specific protease, UCH-L3. [Pilon et al., Biotechnol. Prog., 13, 374-379 (1997)]. U.S. Pat. No. 5,595,887 to Coolidge et al. discloses generalized methods of cloning and isolating peptides. U.S. Pat. No. 5,707,826 to Wagner et al. describes an enzymatic method for modification of recombinant polypeptides.
While the specificity of some proteases is understood, not all proteases have been fully characterized and there are factors that can subtly influence the specificity and efficiency of protease activity. Hence, with currently available processes cleavage of different polypeptide substrates with different proteases requires establishment of unique conditions and/or pre- or post-manipulation of the polypeptides to acquire the desired polypeptide or peptide product. Hence, improved and simplified methods for making peptides are needed. In particular, a method for making peptides is needed which can be universally applied irrespective of the amino acid sequence of the peptide.